Search results

Don't you only train [RET] tokens, why do you train all input embeddings ?

Status: Open.

#10 In dhg-wei/MCL;

· Yang-bug-star opened on Mar 21, 2025

where's your stacked ret token mechanism in code ? It seems that only sequential mechanism was implemented

Status: Open.

#9 In dhg-wei/MCL;

· Yang-bug-star opened on Mar 7, 2025

self.clip_tokenizer use "openai/clip-vit-base-patch32"

Status: Open.

#8 In dhg-wei/MCL;

· Pefect96 opened on Nov 22, 2024

About the evalution

Status: Open.

#7 In dhg-wei/MCL;

· Chenyan722 opened on Nov 21, 2024

Asking for FashionIQ Results

Status: Open.

#5 In dhg-wei/MCL;

· hbhalpha opened on Oct 29, 2024

Generate image features

Status: Open.

#4 In dhg-wei/MCL;

· hmchuong opened on Oct 1, 2024

About the code and dataset

Status: Open.

#2 In dhg-wei/MCL;

· Dzq-ming opened on Jul 5, 2024